基于体素值及其梯度值的三维医学体数据分类

Classification for 3D Medical Volume Data based on Voxels Value and their Grads

结合目前PC图形硬件和医学图像的特点,对基于体素值及其梯度值的体数据分类模型进行了改进,提出了一种改进模型的加速算法.保证不透明度随梯度幅度值的增加而增加的前提下,对梯度幅度值进行放大,以避免过小的梯度幅度值最终产生零值不透明度.结合体数据的窗口变换,用户可以选择不同的梯度幂,观察到不同梯度幅度值的组织边界,为体数据的分类带来了灵活性.

With the research on the classification methods and models of volume data, and characteristic of recent PC graphics hardware and medical images, an improved classification model based on voxels value and their grads and the acceleration algorithm is presented. People can choose different grads for different tissues, which improve the flexibility of volume data classification.